Assess downstream exon expression disturbed by a mutation using RNA-Seq

Question

[Nicolas Rosewick]

Hi,

I've a bunch of RNA-Seq data for different samples and I want to assess the effect of specific mutations localized in gene intron. Looking in IGV we can see that either induction or inhibition of downstream exons are observed. The main issue is that each sample possess a unique mutation, thus each sample disturb one different gene. So no biological replicate are available.. I know for each sample the exact position of the mutation, so the exons that might be disturbed.

My idea to assess the effect of these mutations on downstream exon expression:

1. Align (STAR) and count (htseq-count) read count for each exon
2. Use DEXSeq to normalize exon count across the different sample
3. Normalize exon count for exon length : divide each exon count by its exon length
4. For each gene harboring the specific mutation:
   1. For each sample:
      1. compare the exon count upstream and downstream of the mutation with a t-test: t.test(exon_counts_upstream, exon_counts_downstream)
      2. We expect to have a non-significant p-value for sample that do not harbor the mutation, and a significant p-value for the sample harboring the mutation.

But doing that I have one p-value for each gene and sample. How can I test that globally ( so are the different mutations affecting the exons localized downstream (up or down regulation) ) ?

Thanks